CB1 Cannabinoid Receptors Stimulate Gβγ-GRK2-Mediated FAK Phosphorylation at Tyrosine 925 to Regulate ERK Activation Involving Neuronal Focal Adhesions

CB1 cannabinoid receptors (CB1) are abundantly expressed in the nervous system where they regulate focal adhesion kinase (FAK) and the mitogen-activated protein kinases (MAPK) extracellular signal-regulated kinase 1 and 2 (ERK1/2). However, the role of CB1-stimulated FAK 925 tyrosine phosphorylation (Tyr-P) in regulating ERK1/2 activation remains undefined. Here, immunoblotting analyses using antibodies against FAK phospho-Tyr 925 and ERK2 phospho-Tyr 204 demonstrated CB1-stimulated FAK 925 Tyr-P and ERK2 204 Tyr-P (0–5 min) which was followed by a decline in Tyr-P (5–20 min). CB1 stimulated FAK-Grb2 association and Ras-mediated ERK2 activation. The FAK inhibitors Y11 and PF 573228 abolished FAK 925 Tyr-P and partially inhibited ERK2 204 Tyr-P. FAK 925 Tyr-P and ERK2 204 Tyr-P were adhesion-dependent, required an intact actin cytoskeleton, and were mediated by integrins, Flk-1 vascular endothelial growth factor receptors, and epidermal growth factor receptors. FAK 925 Tyr-P and ERK2 204 Tyr-P were blocked by the Gβγ inhibitor gallein, a GRK2 inhibitor, and GRK2 siRNA silencing, suggesting Gβγ and GRK2 participate in FAK-mediated ERK2 activation. Together, these studies indicate FAK 925 Tyr-P occurs concurrently with CB1-stimulated ERK2 activation and requires the actin cytoskeleton and Gi/oβγ-GRK2-mediated cross-talk between CB1, integrins, and receptor tyrosine kinases (RTKs).


INTRODUCTION
Endocannabinoid signaling in neuronal cells includes a rapid activation of focal adhesion kinase (FAK; Derkinderen et al., 2003;Dalton et al., 2013). FAK is a highly conserved non-receptor protein Tyr kinase that functions as a signal-transducing scaffold protein that regulates multiple cellular functions including proliferation, apoptosis, organization of the actin cytoskeleton, migration, and adhesion (Peng and Guan, 2011). FAK is typically located in focal adhesions where integrin clusters link proteins of the extracellular matrix (ECM) to the actin cytoskeleton. FAK binds to the adaptor protein paxillin to dock with components of the actin cytoskeleton at focal adhesion sites (Parsons, 2003). FAK activation is regulated by Tyr-P and occurs in response to integrin engagement, as well as RTK and G protein-coupled receptor (GPCR) stimulation (Parsons, 2003;Peng and Guan, 2011;Dalton et al., 2013). Tyr 397 is the FAK autophosphorylation site and phosphorylation at this site triggers FAK initial activation. Src family kinases (Src) bind to FAK phospho-Tyr 397 and phosphorylate FAK on additional Tyr residues (Tyr 407, Tyr 576/577, Tyr 861, Tyr 925) that mediate specific FAK functions such as FAK maximal activation and FAK-mediated activation of Ras signal transduction (Schlaepfer et al., 1994;Calalb et al., 1995Calalb et al., , 1996Schlaepfer and Hunter, 1996).

Cell Culture
N18TG2 neuroblastoma cells (passage numbers 25-50) were maintained at 37 • C under a 5% CO 2 atmosphere in Dulbecco's Modified Eagle's Medium (DMEM): Ham's F-12 (1:1) complete with GlutaMax, sodium bicarbonate, and pyridoxine-HCl, supplemented with penicillin (100 units/ml) and streptomycin (100 µg/ml; Gibco Life Technologies, Gaithersburg, MD, USA) and 10% heat-inactivated bovine serum (JRH Biosciences, Lenexa, KS, USA). An aliquot of cannabinoid drug stocks (stored at −20 • C as 10 mM solutions in ethanol) or ethanol (control) was air-dried under sterile conditions in trimethylsilyl-coated glass test tubes and taken up in 100 volumes of 5 mg/ml fatty acid-free bovine serum albumin (BSA) and serially diluted before being added to cells. Where indicated, N18TG2 cells were pretreated with receptor antagonists or other inhibitors before the addition of CB 1 agonists. Pertussis toxin (List Biological Laboratories, Campbell, CA, USA) was added to cells (100 ng/ml) for 16-20 h before addition of agonists.
EDTA, pH 7.4). Cells were resuspended for 20 min on ice in an NP-40 lysis buffer that contained 10 mM NaHEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 100 µM EDTA, 250 µM Na orthovanadate, 1 mM Na fluoride, 1% NP-40, 1 µM DTT, and a protease inhibitor cocktail (EMD Millipore) with broad specificity for the inhibition of aspartic, cysteine, and serine proteases as well as aminopeptidases. Lysates were clarified by centrifugation at 20,000 g at 4 • C and supernatants were stored at −80 • C. Protein concentrations were determined using the Bradford method with BSA as the standard (Bradford, 1976). Lysates were taken up in Laemmli's sample buffer (62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.002% bromophenol blue, 100 mM DTT) and heated at 95 • C for 5 min. Cell lysates were resolved by SDS-PAGE. Gels were pre-equilibrated in Towbin buffer (25 mM Tris Base, 192 mM glycine, 20% methanol, pH 8.3) for 30 min and proteins were transferred to nitrocellulose membranes using a BioRad Trans-Blot Cell. Blots were rinsed one time with Tris-buffered saline (TBS, 20 mM Tris-HCl, 137 mM NaCl, pH 7.4), blocked with Odyssey Blocking buffer, and then incubated with primary antibodies overnight at 4 • C. Blots were washed four times with TBST (TBS containing 0.1% Tween-20), incubated with IRDye 800 CW goat anti-rabbit or IRDye 680 CW goat anti-mouse secondary antibodies (1:15,000) for 1 h at room temperature, followed by three washes with TBST and one wash with TBS. Immunoblots were imaged and bands were quantified by densitometry using Odyssey Infrared Imaging System software (LI-COR Biosciences, Lincoln, NE, USA).

Fluorescence Microscopy Assay
N18TG2 cells were grown on glass coverslips. At 80% confluency, cells were treated with WIN55212-2 (1-1,000 nM, 20 min), latrunculin A (0.01 µg/ml, 30 min), or jasplakinolide (4 nM, 10 min). Following indicated drug treatments, cells were washed with PBS and fixed in 4% paraformaldehyde in cytoskeleton buffer (0.32 M sucrose, 10 mM (2-(4-morpholino)-ethane sulfonic acid, 3 mM MgCl 2 , 138 mM KCl, 2 mM EGTA, pH 6.1). Cells were then permeabilized with 0.5% Triton X-100, blocked with PBS/2% BSA, and treated with 0.1% sodium borohydride. Cells were washed in PBS, then stained with Alexa Fluor 488 phalloidin (200 units/ml) and Texas Red DNase I (9 µg/ml) to evaluate the relative amounts of F-actin and G-actin as previously described (Knowles and McCulloch, 1992). The fluorescence staining procedure was modified by using newer and brighter fluorochromes bound to phalloidin and DNAse I which label the F-actin and G-actin in cells, respectively. Following a wash with PBS, coverslips were mounted onto glass slides with Prolong Antifade reagent. Slides were viewed using a Nikon Eclipse E600 fluorescence microscope (Nikon Instruments, Inc., Melville, NY, USA) and images were digitalized using identical exposure time and brightness settings for all conditions. Quantification of fluorescence was performed using Image-Pro Plus 4.5 Software (Media Cybernetics, Inc., Rockville, MD, USA). Ratios of green (Alexa Fluor 488 phalloidin) to red (Texas Red DNase I) fluorescence were tabulated from three images of 15 or more cells per field. The excitation and emission wavelengths were: 495 nm/518 nm (Alexa Fluor 488 phalloidin) and 597 nm/615 nm (Texas Red DNase I).

GRK2 RNA-Mediated Interference
N18TG2 cells were transfected with 100 nM GRK2-specific siRNA (mouse) or negative control siRNA-A using siRNA transfection reagent according to the manufacturer's protocol (Santa Cruz Biotechnology). Negative control siRNA is a non-targeting siRNA that was used to verify the accuracy of GRK2-specific siRNA and was included in every siRNA experiment. In brief, N18TG2 cells (2 × 10 5 cells/well) were plated on BD Falcon 6-well multiwell plates 24 h before transfection. Cells were then transfected with no siRNA (mock transfection), GRK2-specific siRNA (100 nM), or negative control siRNA (100 nM) for 6 h at 37 • C, 5% CO 2 . Following transfections, cells were cultured in normal growth medium for 48 h and then serum-starved for 24 h before incubation with or without CB 1 agonists. Following drug treatments, cells were harvested with PBS-EDTA, lysed, and protein modification was quantitated by immunoblotting experiments.

Grb2 Co-immunoprecipitation With FAK
N18TG2 cells at 90% confluency were serum-starved (20-24 h) and pre-incubated with THL (1 µM, 2 h) before treatment with CB 1 agonists. Following drug treatments, cells were harvested in PBS-EDTA, and pellets were lysed on ice with NP-40 lysis buffer plus protease inhibitor cocktail for 20 min. Lysates were clarified by centrifugation at 13,500 g at 4 • C for 5 min and protein concentrations were determined (Bradford, 1976). Proteins (500 µg) were immunoprecipitated at 4 • C with antibodies specific for FAK and collected with protein A/G PLUS-Agarose (Santa Cruz Biotechnology). The immune complexes were precipitated by centrifugation at 13,500 g for 5 min at 4 • C, washed three times with ice-cold NP-40 buffer, and boiled in Laemmli's sample buffer. Following centrifugation at 13,500 g for 5 min at 4 • C, supernatants were collected and resolved by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes and immunoblots were stained with anti-FAK or anti-Grb2 primary antibodies. Immunoblots were imaged and bands were quantified by densitometry using Odyssey Infrared Imaging System software.

Statistical Analysis
Graphs and statistical analyses were generated using GraphPad Prism V software (La Jolla, CA, USA). The means between two independent samples were compared using the unpaired Student's t-test. The one way ANOVA test followed by Dunnett's multiple comparison posthoc test was used to compare the means between three or more independent samples.

CB 1 Stimulates Ras-Dependent Activation of the Raf/MEK/ERK Cascade in N18TG2 Cells
The N18TG2 neuroblastoma cell has been used as a model system to study signal transduction pathways regulated by the CB 1 cannabinoid receptor because N18TG2 cells express mRNA and protein for CB 1 but not CB 2 receptors (Mukhopadhyay et al., 2006;Jones et al., 2008). Kinetic analysis revealed the synthetic CB 1 agonist WIN55212-2 (0.01 µM) produced an increase in ERK2 204 Tyr-P in N18TG2 cells that reached maximal levels in 2-5 min and declined to baseline by 10 min (Figures 1A,B). ERK1 (phospho-p44 MAPK) and ERK2 (phospho-p42 MAPK) exhibited similar activation kinetics in N18TG2 cells (data not shown). Consequently, this report focused on the cellular mechanisms that control the level of ERK2 204 Tyr-P. The guanine nucleotide exchange factor Sos1 binds Grb2 at the plasma membrane where Sos1 activates Ras and the downstream Raf/MEK/ERK cascade (Roskoski, 2012). Phosphorylation of FAK at Tyr 925 promotes Grb2 binding to FAK and leads to FAK-mediated activation of the Ras-dependent MAPK pathway through its association with the Grb2/Sos1 complex (Schlaepfer et al., 1994;Schlaepfer and Hunter, 1996). Coimmunoprecipitation experiments confirmed a basal constitutive association of FAK with Grb2 in N18TG2 cells that increased following treatment of the cells with WIN55212-2 (0.01 µM; 1 min 9.8% ± 3.24 increase; 2 min 55.7% ± 17.38 increase; Figure 1C). To determine whether Ras signaling is necessary for ERK2 activation, we examined the effect of the Ras inhibitor FTA on WIN55212-2-stimulated ERK2 204 Tyr-P in N18TG2 cells (Haklai et al., 1998). As shown in Figures 1D-F, pretreatment of N18TG2 cells with FTA (10 µM) partially inhibited WIN55212-2-stimulated ERK2 204 Tyr-P (1 min 53.0% ± 3.40 inhibition; 2 min 49.0% ± 6.30 inhibition). Basal ERK2 204 Tyr-P was not altered by FTA or the vehicle for FTA ( Figure 1F). These findings indicate CB 1 activates the Grb2-Sos1-Ras axis to regulate ERK2 Tyr-P in N18TG2 cells.
CB 1 -Stimulated FAK 925 Tyr-P and ERK2 204 Tyr-P Requires an Intact Actin Cytoskeleton in N18TG2 Cells Integrins are cell adhesion molecules that link the actin cytoskeleton to proteins of the ECM and form the architectural backbone of focal adhesions. Focal adhesions are present in N18TG2 cells as confirmed by anti-vinculin staining, and localization of FAK phosphorylated at Tyr 925 in response to CB 1 stimulation in these structures (unpublished data from our laboratory). Actin exists as a dynamic equilibrium mixture of two forms: polymeric filamentous actin (F-actin) and monomeric globular actin (G-actin; Coumans et al., 1997). The F-actin to G-actin ratio serves as a marker of actin cytoskeleton homeostasis. Studies were conducted to confirm that remodeling of the actin cytoskeleton could be induced in N18TG2 cells using jasplakinolide and latrunculin A. Jasplakinolide is a potent inducer of actin polymerization by stimulating actin filament nucleation, while latrunculin A disrupts the actin cytoskeleton by binding monomeric G-actin and preventing actin polymerization (Spector et al., 1983;Coué et al., 1987;Allingham et al., 2006). In N18TG2 cells, jasplakinolide (4 nM) increased the F-actin to G-actin ratio suggesting actin polymerization, while latrunculin A (0.01 µg/ml) diminished the F-actin to G-actin ratio suggesting actin depolymerization (Figures 3A,B). Cells treated with jasplakinolide also had numerous prominent neurites, while latrunculin A-treated cells had few small processes (Figures 3A,B). WIN55212-2 increased the F-actin to G-actin ratio in a dose-dependent manner in N18TG2 cells ( Figure 3C).

CB 1 -Stimulated FAK 925 Tyr-P and ERK2 204 Tyr-P Are Mediated by Integrins and Receptor Tyrosine Kinases in N18TG2 Cells
Immunoblotting analysis revealed the CB 1 antagonist SR141716A (1 µM) blocked WIN55212-2-stimulated FAK 925 Tyr-P (1 min 83.5% ± 11.40 inhibition; 2 min 82.4% ± 1.34 inhibition) and ERK2 204 Tyr-P (1 min 89.8% ± 5.39 inhibition; 2 min 93.4% ± 4.07 inhibition) which indicates these are CB 1 -dependent events in N18TG2 cells (Figures 4A-D). Previous studies in our laboratory have demonstrated that CB 1 engages in cross-talk with fibronectinbinding integrins, as well as Flk-1 VEGFRs and EGFRs to stimulate maximal FAK activation in N18TG2 cells (Dalton et al., 2013). In data not shown, CB 1 failed to stimulate FAK 925 Tyr-P in the absence of integrin activation in suspended N18TG2 cells, while ERK2 204 Tyr-P was reduced by approximately 50%. These findings led us to hypothesize that both FAK 925 Tyr-P and ERK2 Tyr-P are adhesion-dependent and involve cross-talk between CB 1 and integrins.

DISCUSSION
In N18TG2 neuronal cells, FAK activation begins with FAK 397 Tyr-P, which generates a binding site for Src, resulting in Src activation and Src subsequent phosphorylation of FAK on Tyr 925 (Schlaepfer et al., 1994;Schlaepfer and Hunter, 1996;Parsons, 2003). FAK 925 Tyr-P creates a Src-homology-2-binding site for the adaptor protein Grb2 which links FAK to Grb2-Sos1-mediated activation of the Ras/Raf/MEK/ERK1/2 cascade (Schlaepfer et al., 1994;Schlaepfer and Hunter, 1996;Mitra et al., 2006;Cheung et al., 2008;Roskoski, 2012). CB 1 -stimulated FAK 925 Tyr-P was rapid (0-5 min) and transient, with a more prolonged decline (5-20 min). Like FAK, ERK2 activation is regulated by reversible protein phosphorylation and is dependent on Tyr 204 phosphorylation in the ERK2 activation loop with similar activation kinetics to those of FAK (Roskoski, 2012). The FAK inhibitors Y11 and PF 573228 abolished CB 1 -stimulated FAK 925 Tyr-P, and partially blocked CB 1 -stimulated ERK2 204 Tyr-P. We propose that FAK 925 Tyr-P is a factor in CB 1 -stimulated ERK2 activation in N18TG2 cells. Evidence supporting this includes the Src-dependence of CB 1 -stimulated FAK 925 Tyr-P coupled with our previous demonstration that the Src inhibitor (1 µM, SR), Flk-1 VEGFR antagonist SU 5416 (1 µM), or EGFR antagonist AG 1478 (2 µM) before treatment with 0.01 µM WIN55212-2 (WIN) for 1 or 2 m at 37 • C. Cell lysates were analyzed using western blots and representative blot images are shown. Data are reported as mean ± SEM of (B) % of vehicle-treated FAK 925 Tyr-P (normalized to total FAK levels) at the same time point, (C) % of vehicle-treated ERK2 204 Tyr-P (normalized to total ERK2 levels) at the same time point, and (D) % of basal/time 0 ERK2 204 Tyr-P (normalized to total ERK2 levels). Significance was assessed using One Way ANOVA followed by Dunnett's multiple comparisons posthoc test [ # p < 0.05, *p < 0.01 indicates significantly different from vehicle-treated (at the same time point); **p < 0.05 indicates significantly different from basal/time 0]. For each dataset, cells were cultured and experiments were completed on at least three separate occasions.
The murine neuroblastoma N18TG2 cell is a cell line of neuronal origin. Previous studies have demonstrated that the neuronal isoform of FAK, FAK+6,7, is the major FAK isoform found in hippocampal slices but not astrocytes or microglia (Derkinderen et al., 1996(Derkinderen et al., , 2001. The endocannabinoids anandamide and 2-AG both stimulated FAK+6,7 397 Tyr-P in hippocampal slices that was followed by FAK+6,7 925 Tyr-P ( (Derkinderen et al., 2001). Moreover, endocannabinoidstimulated FAK+6,7 925 Tyr-P was dependent specifically on the Src kinase Fyn in hippocampal slices (Derkinderen et al., 2001). Our work in N18TG2 cells is consistent with what has been shown in hippocampal slices regarding how CB 1 stimulates FAK 925 Tyr-P. We have shown FAK 397 Tyr-P precedes FAK 925 Tyr-P, while work in our lab has also confirmed that FAK 925 Tyr-P is Src-dependent in N18TG2 cells. At present, we do not know the specific FAK isoform expressed in N18TG2 cells and we have not confirmed that FAK 925 Tyr-P is mediated specifically by Fyn. Thus, it is possible the FAK-dependent mechanism linking CB 1 to ERK activation is different in N18TG2 cells compared to hippocampal slices and involves a different FAK isoform and a different Src family member. Although our pathway leading from CB 1 to FAK 925 Tyr-P in the cloned N18TG2 neuronal cell line is similar to what has been observed in hippocampal slices, future studies would need to be done in hippocampal slices to examine the effects of FAK inhibition on CB 1 -stimulated ERK activation to determine if the mechanism is similar to what we report herein.
Studies suggest that actin stress fibers play a significant role in protein recruitment to focal adhesions (Oakes et al., 2012). For this reason, the Tyr-P and activation of FAK are critically dependent on the integrity of the actin cytoskeleton. Disruption of the actin cytoskeleton has been shown to inhibit focal adhesion formation, block Src and FAK translocation to focal adhesions, and block FAK Tyr-P (Lipfert et al., 1992;Fincham et al., 1996;Chen et al., 2000). We confirmed herein that the integrity of the actin cytoskeleton plays an important role in CB 1 -stimulated FAK 925 Tyr-P and ERK2 204 Tyr-P in N18TG2 cells, as treatment with the actin-disrupting drugs cytochalasin D and latrunculin A blocked these events. Data are reported as mean ± SEM of (B) % of vehicle-treated FAK 925 Tyr-P (normalized to total FAK levels) at the same time point, (C) % of vehicle-treated ERK2 204 Tyr-P (normalized to total ERK2 levels) at the same time point, and (D) % of basal/time 0 ERK2 204 Tyr-P (normalized to total ERK2 levels). Significance was assessed using One Way ANOVA followed by Dunnett's multiple comparisons posthoc test [*p < 0.01, # p < 0.05 indicates significantly different from vehicle-treated (at the same time point); **p < 0.05 indicates significantly different from basal/time 0]. (E-H) Cells (2 × 10 5 ) were transfected with no siRNA (mock transfection), GRK2-specific siRNA (100 nM), or negative control siRNA (100 nM) before treatment with 0.01 µM WIN for 2 min at 37 • C. Immunoblot analysis was performed and data are reported as mean ± SEM of the % change over basal (G) FAK 925 Tyr-P (normalized to total FAK levels) and (H) ERK2 204 Tyr-P (normalized to total ERK2 levels). Significance was assessed using One Way ANOVA followed by Dunnett's multiple comparisons posthoc test (*p < 0.01, # p < 0.05 indicates significantly different from GRK2 siRNA at the same time point). For each dataset, cells were cultured and experiments were completed on at least three separate occasions.
FAK is a Tyr kinase that localizes to focal adhesions in which integrins link the actin cytoskeleton to proteins of the ECM. Focal adhesion signaling is mediated by integrins and we have demonstrated that fibronectin (α5β1) and laminin (α6β1, α7β1) integrin receptors are present in N18TG2 cells (Dalton et al., 2013). Integrin activation involves a combination of integrin clustering, as well as ligand occupancy/activation of integrins which mediates cell adhesion to the ECM (Parsons, 2003). α5β1 fibronectin receptors bind to an RGD sequence located in the ECM protein fibronectin, and synthetic peptides, such as RGDS, are effective RGD-binding integrin antagonists because they occupy these integrin sites (Matsuno et al., 1994). Our demonstration that the RGDS peptide significantly reduced both FAK 925 Tyr-P and ERK2 204 Tyr-P in adherent N18TG2 cells attached to their own ECM supports the conclusion that CB 1 engages in interreceptor cross-talk with integrins, as observed for CB 1 in the brain and several other GPCRs (Slack, 1998;Karanian et al., 2005;Teoh et al., 2012;Wang et al., 2018). Furthermore, studies in hippocampal slice cultures demonstrated an integrin antagonist blocks CB 1 -stimulated FAK and ERK activation (Karanian et al., 2005). In our previous work we demonstrated that CB 1 transactivates multiple integrins to stimulate maximal FAK activation in N18TG2 cells (Dalton et al., 2013) but whether GPCRs stimulate integrin activation directly by Gα subunits remains a subject of investigation (Gong et al., 2010).
FAK is an important signaling effector for RTKs and Flk-1 VEGFR and EGFR ligands stimulated FAK Tyr-P (Abedi and Zachary, 1997;Rousseau et al., 2000;Long et al., 2010). We confirmed previous studies that indicated CB 1 transactivates Flk-1 VEGFRs and EGFRs to stimulate ERK1/2 activation (Dalton and Howlett, 2012) and maximal FAK catalytic activity in N18TG2 cells (Dalton et al., 2013), as we demonstrated that inhibition of Flk-1 VEGFRs and EGFRs reduced both. Our data suggest that cooperative signaling between integrins and these RTKs may be involved in CB 1 -stimulated FAK 925 Tyr-P, which would add another interesting dimension to how CB 1 regulates mitogenic signaling in neuronal cells. However, integrins and RTKs can also stimulate Ras-mediated activation of the MAPK cascade independently of FAK and do so by binding Grb2-Sos1 directly or indirectly via a Shc protein (van der Geer and Pawson, 1995;Rojas et al., 1996;Wary et al., 1996;Kroll and Waltenberger, 1997;McKay and Morrison, 2007). Based on the partial dependence of CB 1 -stimulated ERK2 204 Tyr-P on FAK 925 Tyr-P, it is clear that a component of CB 1 -mediated ERK2 signaling is accomplished autonomously from FAK in N18TG2 cells.
Inhibition of FAK 925 Tyr-P and ERK2 204 Tyr-P with pertussis toxin confirmed the involvement of G i/o -coupled CB 1 receptors in these processes, and further evidence indicated that this is mediated by G i/o βγ subunits in N18TG2 cells (Dalton and Howlett, 2012). The Gβγ inhibitor gallein significantly reduced FAK 925 Tyr-P and ERK2 204 Tyr-P demonstrating that FAK and ERK are Gβγ effectors in N18TG2 cells. In addition to Gβγ, GPCRs are regulated by GRKs which participate with β-arrestins in the desensitization-dependent phosphorylation and transient internalization of GPCRs to terminate GPCR-mediated responses (Pitcher et al., 1998;Jin et al., 1999;Kouznetsova et al., 2002;Breivogel et al., 2008). GRK2 is a Gβγ-dependent kinase that has been shown to modulate CB 1 -mediated signaling events (Kouznetsova et al., 2002). We found that a GRK2 inhibitor as well as siRNA silencing of GRK2 expression reduced FAK 925 Tyr-P and ERK2 204 Tyr-P. Information has now emerged that GRK2 can trigger signal propagation from GPCRs and participate in cross-talk between integrins and sphingosine-1-phosphate receptors (Penela et al., 2008). These studies propose that GRK2 forms a complex with G i/o βγ at the plasma membrane where it acts as a scaffold to recruit signaling proteins, such as GIT1, that mediate signals emanating from integrin/sphingosine-1-phosphate stimulation to ultimately promote ERK1/2 activation, focal adhesion turnover, modification of focal adhesion-associated proteins, and cell migration (Penela et al., 2008). Based on this model and our findings with gallein, we speculate that G i/o βγ-bound GRK2 acts in a kinase-independent manner to mediate CB 1 -stimulated FAK and ERK2 activation in N18TG2 cells. Future research is needed to understand the roles GRK2 and G i/o βγ play in this process. It will also be interesting to determine if other GRKs that have been implicated in CB 1 desensitization behave like GRK2 or if these GRKs mediate the decline in CB 1 -stimulated FAK and ERK2 activation in N18TG2 cells (Jin et al., 1999). Recent evidence that GRK5 is involved in CB 2 -induced upregulation and increased activity of 5-HT 2A receptors in neuronal cells also invites opportunities for further investigation (Franklin and Carrasco, 2013).

CONCLUSION
In summary, our data identify novel functions for FAK in CB 1 -stimulated MAPK signaling and characterize an actin cytoskeletal and G i/o βγ-GRK2-mediated signaling pathway utilized by CB 1 to stimulate FAK-mediated MAPK activation. These results provide a mechanism where FAK and MAPK integrate signals from CB 1 , integrins, and RTKs. CB 1 signaling through FAK and MAPK may play important physiological roles in CB 1 regulation of the actin cytoskeleton, neuronal migration, proliferation, and generation of neuritic processes (Harkany et al., 2008;Zorina et al., 2010;Jung et al., 2011;Peng and Guan, 2011;Gaffuri et al., 2012;Roskoski, 2012;Shum et al., 2020).

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.